Processing of satellite images in space-time domain

摘要

Proposed methodology aims at processing of remote sensing images of a single area taken in different times. The suite of images represents a 3D data array structure. Images can be produced combining different bands in order to obtain a space-time representation of environmental parameters as vegetation index, humidity, pollution, temperatures etc. We have experimented calculation of trend polynomials for suites of NDVI values in time of each pixel separately. Polynomial coefficients for the trend of each pixel were used to generate false color images representing the evolution in time of NDVI values for the whole area in study. The work was done in the frame of FP7 project SEE-GRID-SCI, using the application framework CHERS.